SCFIS Europe Rebounds 2% to 2,197.95 as November GRIs Fail to Land

The Analysis

Shanghai Shipping Exchange data put out on 30 October has the SCFIS Europe settlement index at 2,197.95, up 43.18 points, plus 2.0 per cent on the week. Turn to the same week's market notes and nobody's increase is landing: the Gemini Cooperation, the Ocean Alliance and MSC are all quoting early-November boxes at late-October levels. About 63 per cent of Asia-Europe capacity is back through Suez, and Maersk says the rest follows by March 2027. China shipped 86.7 billion dollars of goods to the European Union in September, up 28.94 per cent year on year, and the freight rate still would not move.

A rebound and a dead increase inside one week look like a contradiction. They are not, and knowing which reading runs ahead is worth real money over the next five months.

Take the arithmetic first, because this print is clean. 2,197.95 less 43.18 is 2,154.77, and 43.18 divided by 2,154.77 is 2.004 per cent. The exchange got its own percentage right, which is more than most of the commentary around it manages.

The defect is not in the sum. It is in the tense. This is a settlement index: it aggregates what actually cleared on the previous week's sailings, and those boxes were quoted and booked weeks before they loaded. Assume a booking-to-sailing window of two to three weeks on this lane, replace the assumption with your own, and that 2.0 per cent of 30 October describes the market of roughly the first half of October. Carriers tabled their November increases in the last fortnight of October and watched them die in the first days of November. The two facts are sequential, not contradictory. Read the same week in the wrong order and you get rates recovering; read it in the right order and you get rates starting to roll.

None of this is exotic. Every time-driven index on this lane settles history and publishes it as news. What you want is the leading series, and there are exactly two available to you free, on a weekly cycle, right now.

One is the alliance quote sheet. GEMINI, the Ocean Alliance and MSC all told the market the same thing this week: early November at late October money. Three independent pricing desks arriving at the same number is not a coincidence, it is a clearing price, and it is the least ambiguous signal in the whole story. A general rate increase that does not land is the strongest possible evidence about the balance of power on the lane, because carriers do not file increases they expect to win; they file them to test, and this week the test failed on all three desks.

Watch what those three desks do next, because partial implementation is how a base gets reset without anybody admitting it. When a filing of 400 is settled at 150 in practice, the difference never appears as a rate cut. It appears as a rebate, a loyalty discount, or a spot deal quietly given to whoever asked. Six weeks of that and your base has moved 250 with not one published number changing. So do not score next month by whether another increase gets announced. Score it by what actually cleared: pull every settlement across your own accounts, take the median rather than the lowest, and measure it against the last announcement rather than against last month's average. The distance between announced and cleared is the honest measure of this lane right now, and nobody publishes it.

Size what that failure is worth. Assume the November filings were pitched somewhere between 300 and 500 dollars a 40 foot box, which is the usual shape and is my assumption rather than anything published. On 60 boxes a month that is 18,000 to 30,000 dollars a month you have not paid. Multiply by the months and it dwarfs whatever the index does this quarter. GRIs are the single largest discrete event in most shippers' annual freight cost, and almost nobody books a line item for them.

The other leading series is the blank sailing programme. When carriers release tonnage because a shorter rotation means fewer ships cover the same weekly departure, that capacity only reaches your rate if it stays in the market. If it goes straight back out of the market as cancelled sailings, nothing changes for you at all. Carriers publish blank programmes three to four weeks ahead. Your own file of them, week by week, is the closest thing you will get to a supply forecast for the lane, and it is assembled from public notices.

Split your own invoice before concluding anything from a headline saying rates are flat. The index settles a defined set of charges, and your all-in carries plenty that never reaches it: fuel adjustment, low sulphur, currency, documentation, terminal handling at both ends, and whatever the emissions line is called on European services this quarter. A lane can be flat on freight and climbing on all-in for months, and nearly every review I have sat in used the freight number. Take one November invoice, strip out every non-freight line, and see what share of the total the settlementquiring index actually describes. If that share has been falling, you have been managing the smaller half of your own cost.

Now put the returning tonnage on a calendar, because Maersk gave it an end date and an end date is all you need to turn a story into a schedule. Sixty-three per cent of Asia-Europe capacity has come back through Suez. Assume a rotation round Asia-Europe takes about 84 days via the Cape and about 63 days via Suez, my assumption, check it against your own service strings. A weekly service needs one ship for every seven days of rotation, so the Cape route ties up twelve ships where Suez ties up nine. Three ships out of twelve, or 25 per cent of that string's tonnage, is what a full conversion releases.

At 63 per cent converted, 0.63 times 25 is 15.8 per cent of the tonnage has already come out and gone somewhere. Do not assume it landed on the rates yet, because the ships do not vanish the day the routing changes: they sit idle, they get slow-steamed, they get moved to another trade, and any one of those takes weeks. What is still to come is the remaining 37 per cent, which is 0.37 times 25, or 9.25 per cent of lane tonnage, scheduled to arrive between now and March 2027. Spread evenly over November, December, January, February and March, that is roughly 1.85 per cent of effective supply released per month.

Convert released supply into rate, because that is the bit you actually buy. Take a working elasticity of minus 1.5 to minus 2 on this lane, meaning one per cent of extra capacity takes between 1.5 and 2 per cent off the rate, which is a modelling assumption and not a published constant. Nine point two five per cent of additional effective supply then lands somewhere between minus 14 and minus 18.5 per cent on the lane rate. Applied to 2,197.95: the low end is 2,197.95 times 0.86, which is 1,890 points, and the high end is 2,197.95 times 0.815, which is 1,791 points.

That is the number to argue about. My central expectation, made of nothing but published inputs and stated assumptions, is a European settlement index somewhere between 1,790 and 1,890 points by March 2027, delivered in five monthly instalments rather than in one cliff. If you want to fight the case, fight the assumptions, one at a time, in writing. Fight the rotation days, or the elasticity, or whether released hulls actually become offered slots, or whether carriers absorb it in blank sailings. Those are the four places it can go wrong and every one of them is checkable.

If you want one number showing whether the wave is reaching your lane, use offered slots rather than any price. Add advertised departures on your own services across the next five weeks, multiply by the nominal slots announced for each sailing rather than last year's roster, and keep the running weekly total. Sailings get pulled at short notice, and ship swaps move far more capacity than anybody admits, so this is the only place both surface. My rule: a move of five per cent either side of the five-week average, held two weeks running, counts as a genuine change in supply; anything inside that band is holiday schedules and weather. Time the count on the same weekday every week, because the series is more seasonal than most people assume.

And there is the demand side, which is where this gets interesting, because one of the two headline numbers cannot survive contact with the other. Eighty-six point seven billion dollars of Chinese exports to the European Union in September, up 28.94 per cent. Reverse it: 86.7 divided by 1.2894 is 67.24 billion in September last year, so the increment is 19.46 billion dollars in a single month.

Now ask how many boxes that is. Assume goods worth between 45,000 and 60,000 dollars per 40 foot container, my assumption, adjust it to your own commodity, and 19.46 billion dollars is somewhere between 324,000 and 432,000 incremental boxes in one month. Asia-Europe westbound runs, on my assumption, somewhere around 1.4 to 1.6 million containers a month. Those extra boxes would therefore be between a fifth and nearly a third of the entire lane's monthly volume.

If a fifth to a third of extra cargo had turned up on this lane in September, freight rates would have gone through the roof. They did not. Something in the chain does not reconcile, and there are three candidates. Unit values rose, so the dollars grew faster than the boxes. Or the goods did not move containerised on this lane, going instead by air, by rail, or by sea into somewhere that is not the European Union even though customs records it that way. Or September borrowed October and November.

The third candidate fits the evidence better than the first two, and it fits the failed increases exactly. Demand pulled forward into September is demand missing from October and November, which is precisely when carriers tried to raise prices and could not. It also explains the index: settlements recording early October loadings still carried the September pull-forward volume, and that is why 2,197.95 printed up 2.0 per cent this week while the desks were quietly giving November away.

You can settle it with one number. China's October export print for the European Union, due in roughly three weeks from now. My threshold: anything below 80 billion and the pull-forward reading holds, the release-of-capacity forecast stands, and you should be shortening duration, not lengthening it. Anything above 85 billion and this whole bear case was built on a seasonal wobble, and you should fix longer while you still can. Nothing between those two changes anything, which is why it matters that you have picked the numbers before the data lands rather than after.

One caution on borrowing Maersk's date as everybody's date. A full-return commitment is a statement about a fleet plan, not about the Red Sea, and it assumes nothing else moves in between. Use it as the upper bound of a calendar rather than as a forecast. If you need a working assumption, take the March date, move it two months earlier for the routing decision and two months later for the rate effect, because ships convert before capacity is offered and offered capacity needs another cycle before it reaches what your broker quotes you. Then check both against the October export print and against the blank programme, which are the two things capable of proving it wrong in either direction.

Grace Yang makes this point from the buyer's side better than I do: shipping cost follows volume, and the party that can read the volume first sets the terms. The reason I insist on customs data rather than sentiment is that customs data has a date on it.

Now the part no headline printed, and it is the one that changes what you do on Monday. Everyone is reading the settlement index as the price of the market and it is not; it is the receipt for the market you already bought. The two series that lead it are published earlier in exactly the same week, and almost nobody reads them in order. Thursday's alliance quote sheets tell you where the clearing price is. Blank sailing notices published three to four weeks ahead tell you how much supply will actually be offered. Only then does the settlement print arrive, two to three weeks behind the bookings it measures. Read in that order you get five to seven weeks of lead on the index, every week, for the cost of keeping one file. Read in the reverse order, which is how these things are normally consumed, you get a two-week delay and a confident opinion formed after the fact.

Work the decision at your own volume before you touch the contract. Assume 60 boxes a month from Shanghai to Hamburg, and assume your current base freight is 2,000 dollars a box, which is your number to supply, not mine. Flat for twelve months that is 120,000 dollars a month and 1,440,000 dollars a year. Put my path against it: an average of about 12 per cent below today across the year, because the decline spreads over five months and then settles, gives 1,267,200 dollars. The gap is 172,800 dollars in favour of floating. Reverse it, assume instead that the return stalls and rates average 10 per cent above today, and floating costs you 144,000 dollars. There is your asymmetry, stated with its assumptions attached: roughly 173,000 of probable benefit against roughly 144,000 of improbable loss, which is why the answer here is not fixed twelve months and it is not fully floating either.

What you want instead is duration you control. Sign three-month tranches with a repricer that fires if the lane moves 8 per cent either way, so your downside is bounded and you still collect the drip. Better, if your lane is quoted off a published index, ask for a collar: a floor set at 5 per cent below today's level and a ceiling at today plus 3 per cent, referenced to a named weekly print on a named date. Carriers accept collars more readily than they accept low fixed numbers, because the collar leaves them the top half of the distribution and hand you the part you cannot predict. If none of that is available to you, take the three-month deal and re-tender on the last Thursday of every month, whether or not you think you need to.

There is a version of this that goes badly for anyone who moves too far. If the release never arrives because Red Sea risk steps back up and the return pauses below current levels, the 9.25 per cent never lands and the index sits between 2,100 and 2,200 into next spring, and the floating camp pays for nothing. If the return accelerates and completes by January rather than March, the monthly drip roughly doubles to around 3.7 per cent of supply a month and you should expect the low end of the range sooner. And if carriers decide to neutralise it, they can: a weekly programme with a tenth of sailings cancelled absorbs the whole nine point two five per cent on its own, which is exactly why blank notices belong on your desk four weeks early rather than in your inbox the week the ship disappears.

Two percent is not a rebound. It is a receipt for freight agreed before anyone tried to raise prices, arriving the same week that every pricing desk on the lane quietly admitted it could not.

↑ Back to the key points

— By Vivian Zhao